A new fork for clinical application: Targeting forkhead transcription factors in cancer therapy

RAS-ERK-MAPK (Mitogen-activated protein kinase) pathway plays an essential role in proliferation, differentiation, and tumor progression. In this study, we showed that ERK downregulated FOXO3a through directly interacting with and phosphorylating FOXO3a at Serine 294, Serine 344, and Serine 425. ERK-phosphorylated FOXO3a was degraded by MDM2-mediated ubiquitin-proteosome pathway. FOXO3a phosphorylation and degradation consequently promoted cell proliferation and tumorigenesis. However, the non-phosphorylated FOXO3a mutant, which was resistant to the interaction and degradation by MDM2, resulted in inhibition of tumor formation. Forkhead O transcription factors (FOXOs) are important in the regulation of cellular functions including cell cycle arrest and cell death. Perturbation of FOXOs function leads to deregulated cell proliferation and cancer. Inactivation of FOXO proteins by activation of cell survival pathways, such as PI3K/AKT/IKK, is associated with tumorigenesis. Our study will further highlight FOXOs as new therapeutic targets in a broad spectrum of cancers. ^ Chemotherapeutic drug resistance is the most concerned problem in cancer therapy as resistance ultimately leads to treatment failure of cancer patients. In another study, we showed that blocking ERK activity with AZD6244, an established MEK1/2 inhibitor currently under human cancer clinical trials, enhances FOXO3a expression in various human cancer cell lines in vitro, and also in human colon cancer cell xenografts in vivo. Knocking down FOXO3a and its downstream gene Bim impaired AZD6244-induced growth suppression, whereas restoring activation of FOXO3a sensitized human cancer cell to AZD6244-induced growth arrest and apoptosis. More importantly, AZD6244-resistant cancer cells showed impaired endogenous FOXO3a nuclear translocation, reduced FOXO3a-Bim promoter association and significantly decreased Bim expression in response to AZD6244. AZD6244-resistant cancer cells can be sensitized to API-2 (an AKT inhibitor) and LY294002 (a PI3K inhibitor) in suppressing cell growth and colony formation, these inhibitors were known to enhance FOXO3a activity/nuclear translocation through inhibiting PI3K-AKT pathway. This study reveals novel molecular mechanism contributing to AZD6244-resistance through regulation of FOXO3a activity, further provides significant clinical implication of combining AZD6244 with PI3K/AKT inhibitors for sensitizing AZD6244-resistant cancer cells by activating FOXO3a. FOXO3a activation can be an essential pharmacological target and indicator to mediate and predict AZD6244 efficacy in clinical use. ^

Regulation of estrogen receptor alpha by the tumor suppressor p53 in breast cancer cells

Estrogen receptor (ER) and the tumor suppressor p53 are key prognostic indicators in breast cancer. Estrogen signaling through its receptor (ER) controls proliferation of normal as well as transformed mammary epithelial cells, and the presence of ER is established as a marker of good prognosis and response to therapy. The p53 tumor suppressor gene is often referred to as the "cellular gatekeeper" due to its extensive control of cell proliferation and apoptosis. Loss of functional p53 is a negative prognostic indicator and is correlated with lack of response to antiestrogens, reduced disease-free interval and increased chance of disease recurrence. Clinical studies have demonstrated that tumors with mutated p53 tend to be ER negative, while ER positive tumors tend to have wild type p53. ^ Recent studies from our lab indicate that p53 genotype correlates with estrogen receptor expression in mammary tumors in vivo. We therefore hypothesized that p53 regulates ER expression in mammary cancer cells by recruitment of specific cofactors to the ER promoter. To test this, MCF-7 cells were treated with doxorubicin or ionizing radiation, both of which stimulated significant increases in p53 expression, as expected, but also increased ER expression in a p53-dependent manner. Furthermore, in cells treated with siRNA targeting p53, both p53 and ER protein levels were significantly reduced. P53 was also demonstrated to transcriptionally regulate the ER promoter in luciferase assays and chromatin immunoprecipitation assays showed that p53 was recruited to the ER promoter along with CARM1, CBP, c-Jun and Sp1 and that this multifactor complex was formed in a p53-dependent manner. The regulation of ER by p53 has therapeutic implications, as the treatment of breast cancer cells with doxorubicin sensitized these cells to tamoxifen treatment. Furthermore, response to tamoxifen as well as to estrogen was dependent on p53 expression in ER positive human breast cancer cells. Taken together, these data demonstrate that p53 regulates ER expression through transcriptional control of the ER promoter, accounting for their concordant expression in human breast cancer and identifying potentially beneficial therapeutic strategies for the treatment of ER positive breast cancers. ^

TRIM24-REGULATED ESTROGEN RESPONSE IS DEPENDENT ON SPECIFIC HISTONE MODIFICATIONS IN BREAST CANCER CELLS

In this dissertation, I discovered that function of TRIM24 as a co-activator of ERα-mediated transcriptional activation is dependent on specific histone modifications in tumorigenic human breast cancer-derived MCF7 cells. In the first part, I proved that TRIM24-PHD finger domain, which recognizes unmethylated histone H3 lysine K4 (H3K4me0), is critical for ERα-regulated transcription. Therefore, when LSD1-mediated demethylation of H3K4 is inhibited, activation of TRIM24-regulated ERα target genes is greatly impaired. Importantly, I demonstrated that TRIM24 and LSD1 are cyclically recruited to estrogen responsive elements (EREs) in a time-dependent manner upon estrogen induction, and depletion of their expression exert corresponding time-dependent effect on target gene activation. I also identified that phosphorylation of histone H3 threonine T6 disrupts TRIM24 from binding to the chromatin and from activating ERα-regulated targets. In the second part, I revealed that TRIM24 depletion has additive effect to LSD1 inhibitor- and Tamoxifen-mediated reduction in survival and proliferation in breast cancer cells.

Mechanisms Underlying the Heterogeneous Sensitivities of Cancer Cells to Proteasome Inhibitors

The mechanisms underlying cellular response to proteasome inhibitors have not been clearly elucidated in solid tumor models. Evidence suggests that the ability of a cell to manage the amount of proteotoxic stress following proteasome inhibition dictates survival. In this study using the FDA-approved proteasome inhibitor bortezomib (Velcade®) in solid tumor cells, we demonstrated that perhaps the most critical response to proteasome inhibition is repression of global protein synthesis by phosphorylation of the eukaryotic initiation factor 2-α subunit (eIF2α). In a panel of 10 distinct human pancreatic cancer cells, we showed marked heterogeneity in the ability of cancer cells to induce eIF2α phosphorylation upon stress (eIF2α-P); lack of inducible eIF2α-P led to excessive accumulation of aggregated proteins, reactive oxygen species, and ultimately cell death. In addition, we examined complementary cytoprotective mechanisms involving the activation of the heat shock response (HSR), and found that induction of heat shock protein 70 kDa (Hsp72) protected against proteasome inhibitor-induced cell death in human bladder cancer cells. Finally, investigation of a novel histone deacetylase 6 (HDAC6)-selective inhibitor suggested that the cytoprotective role of the cytoplasmic histone deacetylase 6 (HDAC6) in response to proteasome inhibition may have been previously overestimated.

The effect of protein tyrosine phosphatase SHP1 on tumorigenicity of the MDA -MB231 breast cancer cells

SHP1 is a cytosolic protein tyrosine phosphatase that contains two SH2 domains. It is highly expressed in hematopoietic cells and expressed in normal epithelium at lower levels. While SHP1 in hematopoietic cells is thought to be a negative regulator of cellular signaling by associating with and dephosphorylating various receptors and their downstream effectors after they become activated, its precise function in epithelium remains to be understood. The potential involvement of SHP1 in human tumorigenesis has been hypothesized from the findings that SHP1 can interact with, dephosphorylate, and regulate the activity of several protein tyrosine kinases (PTKs) implicated in human cancer. These PTKs include epidermal growth factor receptor (EGFR) and Src. Such speculation is also supported by the report that SHP1 is overexpressed in human ovarian cancers. ^ Here we report, for the first time, that the levels of SHP1 expression and activity are altered in human breast cancer cells in comparison with normal breast epithelium. In particular, SHP1 expression is nearly lost in the breast cancer cell lines MDA-MB231 and MDA-MB435. After the re-introduction of SHP1 both in wild type (wt) and enzymatically inactive (dn) forms, into the MDA-MB231 cells, we observed no changes in cellular proliferation. However, the overexpression of wt SHP1 led to increased anchorage-independent growth in the MDA-MB231 cells. SHP1 phosphatase activity is essential for such an increase since the overexpression of dn SHP1 had no effect. Enhanced turnorigenicity in nude mice was also observed in the MDA-MB231 cells overexpressing wt SHP1, but not dn SHP1, suggesting the crucial function of SHP1 enzymatic activity in this process. Our observations in this study indicate that SHP1 promotes tumorigenesis by a mechanism or mechanisms apart from enchancing angiogenesis. In addition, we have found no evidence that the overexpression of SHP1 could affect metastatic potential in the MDA-MB231 cells. ^ In the MDA-MB231 cells stably transfected with either wt or dn SHP1 the peak level of EGFR tyrosine phosphorylation induced by EGF, as well as the sensitivity to EGF stimulation, was not altered. However, the overexpression of wt SHP1 led to a slight increase in the kinetics of EGFR dephosphorylation, whereas the overexpression of dn SHP1 led to slightly delayed kinetics of EGFR dephosphorylation. The overexpression of either the wt or dn SHP1 did not lead to any significant increase in Src kinase activity. ^ In NIH3T3 cells, the transient overexpression of SHP1 led to no significant changes in MAP kinase (ERK2) activation by EGF or Akt activation by PDGF. In 3T3H4 cells, the transient overexpression of SHP1 led to no significant changes in MAP kinase (ERK2) activation by heregulin. The transient overexpression of wt SHP1 in the MDA-MB231 cells caused an apparent increase, ranging from 10% to 20%, in the G0/G1 population of the cells with a corresponding decrease in the S phase population. ^ In order to understand the mechanisms by which SHP1 exerts its positive effect on the tumorigenic potential of the MDA-MB231 cells, we employed two-dimensional electrophoresis in an attempt to identify cellular protein(s) with significantly altered tyrosine phosphorylation level upon wt SHP1 overexpression. The overexpression of wt SHP1 but not dn SHP1, leads increased tyrosine phosphorylation of a protein with a molecular weight of approximately 40 kDa and a pI between 5.9 to 6.6. ^

The functional role of the tumor suppressor MMAC /PTEN in glioblastoma multiforme and prostate cancer

Investigations into the molecular basis of glioblastoma multiforme led to the identification of a putative tumor suppressor gene, MMAC/ PTEN. Initial studies implicated MMAC/PTEN in many different tumor types, and identified a protein phosphatase motif in its sequence. This project aimed to identify the biological and biochemical functions of MMAC/PTEN by transiently expressing the gene in cancer cells that lack a functional gene product. ^ Expression of MMAC/PTEN mildly suppressed the growth of U251 human glioma cells and abrogated the growth advantage mediated by overexpression of the epidermal growth factor receptor (EGFR). Immunoblotting demonstrated that MMAC/PTEN expression did not affect the phosphorylation of the EGFR itself, or the intermediates of several downstream signaling pathways. However, MMAC/PTEN expression significantly reduced the phosphorylation and catalytic activity of the proto-oncogene Akt/PKB. While Akt/PKB regulates the survival of many cell types, expression of MMAC/PTEN did not induce apoptosis in adherent U251 cells. Instead, MMAC/PTEN expression sensitized the cells to apoptosis when maintained in suspension (anoikis). As the survival of suspended cells is one of the hallmarks leading to metastasis, MMAC/PTEN expression was examined in a system in which metastasis is more clinically relevant, prostate cancer. ^ Expression of MMAC/PTEN in both LNCaP and PC3-P human prostate cancer cells specifically inhibited Akt/PKB phosphorylation. MMAC/PTEN expression in LNCaP cells resulted in a profound inhibition of growth that was significantly greater than that achieved with expression of p53. Expression of MMAC/PTEN in PC3-P cells resulted in greater growth inhibition than was observed in U251 glioma cells, but less than was observed in LNCaP cells, or upon p53 expression. To determine if MMAC/PTEN could function as a tumor suppressor in vivo, the effects of MMAC/PTEN expression on PC3-P cells implanted orthotopically in nude mice were examined. The ex-vivo expression of MMAC/PTEN did not decrease tumor incidence, but it did significantly decrease tumor size and metastasis. In-vivo expression of MMAC/PTEN in pre-established PC3-P tumors did not significantly inhibit tumor incidence or size, but did inhibit metastasis formation. ^ These studies demonstrate that MMAC/PTEN is a novel and important tumor suppressor gene, which functions to downregulate an important cell survival signaling pathway. ^

Identification of a biomarker panel for colorectal cancer diagnosis

Background Malignancies arising in the large bowel cause the second largest number of deaths from cancer in the Western World. Despite progresses made during the last decades, colorectal cancer remains one of the most frequent and deadly neoplasias in the western countries. Methods A genomic study of human colorectal cancer has been carried out on a total of 31 tumoral samples, corresponding to different stages of the disease, and 33 non-tumoral samples. The study was carried out by hybridisation of the tumour samples against a reference pool of non-tumoral samples using Agilent Human 1A 60-mer oligo microarrays. The results obtained were validated by qRT-PCR. In the subsequent bioinformatics analysis, gene networks by means of Bayesian classifiers, variable selection and bootstrap resampling were built. The consensus among all the induced models produced a hierarchy of dependences and, thus, of variables. Results After an exhaustive process of pre-processing to ensure data quality--lost values imputation, probes quality, data smoothing and intraclass variability filtering--the final dataset comprised a total of 8, 104 probes. Next, a supervised classification approach and data analysis was carried out to obtain the most relevant genes. Two of them are directly involved in cancer progression and in particular in colorectal cancer. Finally, a supervised classifier was induced to classify new unseen samples. Conclusions We have developed a tentative model for the diagnosis of colorectal cancer based on a biomarker panel. Our results indicate that the gene profile described herein can discriminate between non-cancerous and cancerous samples with 94.45% accuracy using different supervised classifiers (AUC values in the range of 0.997 and 0.955)

Identification of a biomarker panel for colorectal cancer diagnosis

Background:Malignancies arising in the large bowel cause the second largest number of deaths from cancer in the Western World. Despite progresses made during the last decades, colorectal cancer remains one of the most frequent and deadly neoplasias in the western countries. Methods: A genomic study of human colorectal cancer has been carried out on a total of 31 tumoral samples, corresponding to different stages of the disease, and 33 non-tumoral samples. The study was carried out by hybridisation of the tumour samples against a reference pool of non-tumoral samples using Agilent Human 1A 60-mer oligo microarrays. The results obtained were validated by qRT-PCR. In the subsequent bioinformatics analysis, gene networks by means of Bayesian classifiers, variable selection and bootstrap resampling were built. The consensus among all the induced models produced a hierarchy of dependences and, thus, of variables. Results: After an exhaustive process of pre-processing to ensure data quality--lost values imputation, probes quality, data smoothing and intraclass variability filtering--the final dataset comprised a total of 8, 104 probes. Next, a supervised classification approach and data analysis was carried out to obtain the most relevant genes. Two of them are directly involved in cancer progression and in particular in colorectal cancer. Finally, a supervised classifier was induced to classify new unseen samples. Conclusions: We have developed a tentative model for the diagnosis of colorectal cancer based on a biomarker panel. Our results indicate that the gene profile described herein can discriminate between non-cancerous and cancerous samples with 94.45% accuracy using different supervised classifiers (AUC values in the range of 0.997 and 0.955).

Suppression of breast cancer growth and metastasis by a serpin myoepithelium-derived serine proteinase inhibitor expressed in the mammary myoepithelial cells

A serpin was identified in normal mammary gland by differential cDNA sequencing. In situ hybridization has detected this serpin exclusively in the myoepithelial cells on the normal and noninvasive mammary epithelial side of the basement membrane and thus was named myoepithelium-derived serine proteinase inhibitor (MEPI). No MEPI expression was detected in the malignant breast carcinomas. MEPI encodes a 405-aa precursor, including an 18-residue secretion signal with a calculated molecular mass of 46 kDa. The predicted sequence of the new protein shares 33% sequence identity and 58% sequence similarity to plasminogen activator inhibitor (PAI)-1 and PAI-2. To determine whether MEPI can modulate the in vivo growth and progression of human breast cancers, we transfected a full-length MEPI cDNA into human breast cancer cells and studied the orthotopic growth of MEPI-transfected vs. control clones in the mammary fat pad of athymic nude mice. Overexpression of MEPI inhibited the invasion of the cells in the in vitro invasion assay. When injected orthotopically into nude mice, the primary tumor volumes, axillary lymph node metastasis, and lung metastasis were significantly inhibited in MEPI-transfected clones as compared with controls. The expression of MEPI in myoepithelial cells may prevent breast cancer malignant progression leading to metastasis.

β-Catenin mutations in cell lines established from human colorectal cancers

β-catenin has functions as both an adhesion and a signaling molecule. Disruption of these functions through mutations of the β-catenin gene (CTNNB1) may be important in the development of colorectal tumors. We examined the entire coding sequence of β-catenin by reverse transcriptase–PCR (RT-PCR) and direct sequencing of 23 human colorectal cancer cell lines from 21 patients. In two cell lines, there was apparent instability of the β-catenin mRNA. Five different mutations (26%) were found in the remaining 21cell lines (from 19 patients). A three-base deletion (codon 45) was identified in the cell line HCT 116, whereas cell lines SW 48, HCA 46, CACO 2, and Colo 201 each contained single-base missense mutations (codons 33, 183, 245, and 287, respectively). All 23 cell lines had full-length β-catenin protein that was detectable by Western blotting and that coprecipitated with E-cadherin. In three of the cell lines with CTNNB1 mutations, complexes of β-catenin with α-catenin and APC were detectable. In SW48 and HCA 46, however, we did not detect complexes of β-catenin protein with α-catenin and APC, respectively. These results show that selection of CTNNB1 mutations occurs in up to 26% of colorectal cancers from which cell lines are derived. In these cases, mutation selection is probably for altered β-catenin function, which may significantly alter intracellular signaling and intercellular adhesion and may serve as a complement to APC mutations in the early stages of tumorigenesis.

Human cells compromised for p53 function exhibit defective global and transcription-coupled nucleotide excision repair, whereas cells compromised for pRb function are defective only in global repair

After exposure to DNA-damaging agents, the p53 tumor suppressor protects against neoplastic transformation by inducing growth arrest and apoptosis. A series of investigations has also demonstrated that, in UV-exposed cells, p53 regulates the removal of DNA photoproducts from the genome overall (global nucleotide excision repair), but does not participate in an overlapping pathway that removes damage specifically from the transcribed strand of active genes (transcription-coupled nucleotide excision repair). Here, the highly sensitive ligation-mediated PCR was employed to quantify, at nucleotide resolution, the repair of UVB-induced cyclobutane pyrimidine dimers (CPDs) in genetically p53-deficient Li–Fraumeni skin fibroblasts, as well as in human lung fibroblasts expressing the human papillomavirus (HPV) E6 oncoprotein that functionally inactivates p53. Lung fibroblasts expressing the HPV E7 gene product, which similarly inactivates the retinoblastoma tumor-suppressor protein (pRb), were also investigated. pRb acts downstream of p53 to mediate G1 arrest, but has no demonstrated role in DNA repair. Relative to normal cells, HPV E6-expressing lung fibroblasts and Li–Fraumeni skin fibroblasts each manifested defective CPD repair along both the transcribed and nontranscribed strands of the p53 and/or c-jun loci. HPV E7-expressing lung fibroblasts also exhibited reduced CPD removal, but only along the nontranscribed strand. Our results provide striking evidence that transcription-coupled repair, in addition to global repair, are p53-dependent in UV-exposed human fibroblasts. Moreover, the observed DNA-repair defect in HPV E7-expressing cells reveals a function for this oncoprotein in HPV-mediated carcinogenesis, and may suggest a role for pRb in global nucleotide excision repair.

Genetic instability of cancer cells is proportional to their degree of aneuploidy

Genetic and phenotypic instability are hallmarks of cancer cells, but their cause is not clear. The leading hypothesis suggests that a poorly defined gene mutation generates genetic instability and that some of many subsequent mutations then cause cancer. Here we investigate the hypothesis that genetic instability of cancer cells is caused by aneuploidy, an abnormal balance of chromosomes. Because symmetrical segregation of chromosomes depends on exactly two copies of mitosis genes, aneuploidy involving chromosomes with mitosis genes will destabilize the karyotype. The hypothesis predicts that the degree of genetic instability should be proportional to the degree of aneuploidy. Thus it should be difficult, if not impossible, to maintain the particular karyotype of a highly aneuploid cancer cell on clonal propagation. This prediction was confirmed with clonal cultures of chemically transformed, aneuploid Chinese hamster embryo cells. It was found that the higher the ploidy factor of a clone, the more unstable was its karyotype. The ploidy factor is the quotient of the modal chromosome number divided by the normal number of the species. Transformed Chinese hamster embryo cells with a ploidy factor of 1.7 were estimated to change their karyotype at a rate of about 3% per generation, compared with 1.8% for cells with a ploidy factor of 0.95. Because the background noise of karyotyping is relatively high, the cells with low ploidy factor may be more stable than our method suggests. The karyotype instability of human colon cancer cell lines, recently analyzed by Lengnauer et al. [Lengnauer, C., Kinzler, K. W. & Vogelstein, B. (1997) Nature (London) 386, 623–627], also corresponds exactly to their degree of aneuploidy. We conclude that aneuploidy is sufficient to explain genetic instability and the resulting karyotypic and phenotypic heterogeneity of cancer cells, independent of gene mutation. Because aneuploidy has also been proposed to cause cancer, our hypothesis offers a common, unique mechanism of altering and simultaneously destabilizing normal cellular phenotypes.

Cytokine-mediated enhancement of epidermal growth factor receptor expression provides an immunological approach to the therapy of pancreatic cancer

The increased expression of epidermal growth factor receptor induced by tumor necrosis factor α renders pancreatic cancer cells more susceptible to antibody-dependent cellular cytotoxicity by a mAb specific for this receptor. Laboratory studies with athymic mice bearing xenografts of human pancreatic cancer cells demonstrated a cytokine-induced ability of the mAb to cause significant tumor regression. In a phase I/II clinical trial, 26 patients with unresectable pancreatic cancer were enrolled into three cohorts receiving variable amounts of the antibody together with a constant amount of tumor necrosis factor α. With increasing doses of antibody, the growth of the primary tumor was significantly inhibited. This was reflected by a longer median survival, with one complete remission lasting for 3 years obtained with the highest dose of antibody employed. Thus, a combination of the cytokine, tumor necrosis factor α, with a mAb to the epidermal growth factor receptor offers a potentially useful approach for the treatment of pancreatic cancer.

Direct Visualization of the Human Estrogen Receptor α Reveals a Role for Ligand in the Nuclear Distribution of the Receptor

The human estrogen receptor α (ER α) has been tagged at its amino terminus with the S65T variant of the green fluorescent protein (GFP), allowing subcellular trafficking and localization to be observed in living cells by fluorescence microscopy. The tagged receptor, GFP-ER, is functional as a ligand-dependent transcription factor, responds to both agonist and antagonist ligands, and can associate with the nuclear matrix. Its cellular localization was analyzed in four human breast cancer epithelial cell lines, two ER+ (MCF7 and T47D) and two ER− (MDA-MB-231 and MDA-MB-435A), under a variety of ligand conditions. In all cell lines, GFP-ER is observed only in the nucleus in the absence of ligand. Upon the addition of agonist or antagonist ligand, a dramatic redistribution of GFP-ER from a reticular to punctate pattern occurs within the nucleus. In addition, the full antagonist ICI 182780 alters the nucleocytoplasmic compartmentalization of the receptor and causes partial accumulation in the cytoplasm in a process requiring continued protein synthesis. GFP-ER localization varies between cells, despite being cultured and treated in a similar manner. Analysis of the nuclear fluorescence intensity for variation in its frequency distribution helped establish localization patterns characteristic of cell line and ligand. During the course of this study, localization of GFP-ER to the nucleolar region is observed for ER− but not ER+ human breast cancer epithelial cell lines. Finally, our work provides a visual description of the “unoccupied” and ligand-bound receptor and is discussed in the context of the role of ligand in modulating receptor activity.

Endogenous, hyperactive Rac3 controls proliferation of breast cancer cells by a p21-activated kinase-dependent pathway

Uncontrolled cell proliferation is a major feature of cancer. Experimental cellular models have implicated some members of the Rho GTPase family in this process. However, direct evidence for active Rho GTPases in tumors or cancer cell lines has never been provided. In this paper, we show that endogenous, hyperactive Rac3 is present in highly proliferative human breast cancer-derived cell lines and tumor tissues. Rac3 activity results from both its distinct subcellular localization at the membrane and altered regulatory factors affecting the guanine nucleotide state of Rac3. Associated with active Rac3 was deregulated, persistent kinase activity of two isoforms of the Rac effector p21-activated kinase (Pak) and of c-Jun N-terminal kinase (JNK). Introducing dominant-negative Rac3 and Pak1 fragments into a breast cancer cell line revealed that active Rac3 drives Pak and JNK kinase activities by two separate pathways. Only the Rac3–Pak pathway was critical for DNA synthesis, independently of JNK. These findings identify Rac3 as a consistently active Rho GTPase in human cancer cells and suggest an important role for Rac3 and Pak in tumor growth.